Optically active actinol is known inter alia as an important building block for the synthesis of carotenoids such as zeaxanthin [Pure & Appl. Chem. 51, 535-564 (1979)]. In known processes actinol is manufactured from the trimethylcyclo-hexadione levodione. One such process is based, for example, on the catalytic hydrogenation of levodione in the presence of Raney nickel with a low base content [Helv. Chim. Acta 59, 1832 (1976)]. The poor diastereoselectivity of the heterogeneous hydrogenation, which leads to a 3-4:1 mixture of trans-and cis-actinol, results, however, in a loss in yield by virtue of the complicated purification; moreover there is a danger of a racemization of the levodione under the reaction conditions used in the heterogeneous catalysis. An alternative purification using a suitable distillation column is described in EP-A 0 775 685, but the yields are modest. The interest in manufacturing methods which yield actinol with high enantiomeric and diastereomeric purity is as great now as it always has been.
Processes for the manufacture of alcohols from ketones by transfer hydrogenation are known. Thus, for example, R. Noyori et al. [Acc. Chem. Res. 30, 97-102 (1997)] have investigated the asymmetric transfer hydrogenation of aryl alkyl ketones, such as, for example, acetophenone, which can be hydrogenated with high optical and chemical yield in the presence of a hydrogen donor, e.g. isopropanol/potassium hydroxide or acetic acid/triethylamine, and a ruthenium complex. However, not only the chemical yields, but also the optical yields diminish considerable when, for example, dialkyl ketones are used as substrates.
Surprisingly, it has now been found that (R)-levodione can be converted into (R,R)-actinol in good chemical yield and in high optical yield by transfer hydrogenation in the presence of an amino-amide-ruthenium complex.